Digital I/O

cancel
Showing results for 
Search instead for 
Did you mean: 

Need info on stability of the NI 6723 analog output voltage

I have a requirement which I plan to use a NI 6723 for. The question has come up as to the voltage stability with time ( for now, assume no temp. changes). A typical voltage for my application is 50 mVDC.  Testing is being run now to see, if any voltage changes occur over a long ( i.e. 1 - 2 hrs) period of time.  I don't see any "stability" info in the 6723 data sheet.  Would anyone suggest where I might find that info?  Or, do I have to call my local NI rep and start that long process?
 
Thanks,
 
Henry Park
US Genomics
0 Kudos
Message 1 of 6
(3,791 Views)
Hi Henry,

The specifications given for the NI 6723 on its product page indicate that the device should operate with an absolute accuracy of 10.78mV over the voltage range of +/-10V.  An important point here though is that the DAC on the card will settle within this +/-10.78mV of the value specified in software if it is given the proper settling time.  For DC voltage generation the settling time should not be an issue because you send a fixed voltage update to the DAC, and since that voltage doesn't change over time the initial settling time for the DAC to reach that value should not be noticeable.  If you are wanting to generate at 50mVDC then an accuracy spec of +/-10.78mV may not be acceptable (results in an error of about 22%).  As an alternative, the PCI-6703 and PCI/PXI-6704 analog output devices all have an absolute accuracy spec of 1mV over the voltage range of +/-10V (2% error when generating at 50mVDC).

Regards,
Andrew W
National Instruments
Message 2 of 6
(3,780 Views)

Hi Andrew,

 

what about the stability in case of analog AC signals??

i have tried generating a sine wave of say 50hz and i do not get a constant frequency output at the PXI 6723 card.

Is this an issue with the card or the DAQmx driver??

0 Kudos
Message 3 of 6
(3,187 Views)

Hi Henry,

 

Any variation in the frequency of your output signal will be a function of the jitter of the sample clock on the board. If you open the Cont Gen Voltage Wfm-Int Clk example (Help»Find Examples»Hardware Input and Output»DAQmx»AnalogGeneration»Voltage) how much variation do you see in the signal frequency? If you have a very precise clock (from a 6653 timing module for example) you could use that as the sample clock on your 6723.

Sean Ferguson
Application Engineering Specialist | RF and Reconfigurable Test
0 Kudos
Message 4 of 6
(3,174 Views)

Hi Sean,

In my case the frequency fluctuations while producing a single Analog Cont waveform is minimal but when i am simultaneously producing say 10 sine waves on the same 6723 card, then any change i make in the frequency of one wave is reflected uniformly in the other nine! what could the reason for this be. Also this behaviour occurs only in the case of decimal point change as in if i change from 50 to 50.2, but the same doen't occur when i am varying it to say 51Hz.

 

 

0 Kudos
Message 5 of 6
(3,162 Views)

Hi ICs,

 

One thing to understand about the 6723 is that there is only one timebase for all of the analog output channels. What this means is that each channel will be sampled (updated) at the same time (See page 3-6 of the manual). Another thing to point out is that only a discreet number of signal frequencies can be generated because the sample clock is an integer divisor of the master timebase. So, for example, if you specify a sample clock rate of 60 Hz, what will actually be produced will be something like 60.002 Hz. You can find out what this value is by using a DAQmx Timing Property node:

 

SampClkRate.png

 

Hope this helps!

 

 

Sean Ferguson
Application Engineering Specialist | RF and Reconfigurable Test
0 Kudos
Message 6 of 6
(3,152 Views)